sign in sign up
<<
Home , GATA1, GATA2, Glucocorticoid receptor

113 Inhibition of GATA2-dependent of the TSHb by ligand-bound a

Koji Nagayama, Shigekazu Sasaki, Akio Matsushita, Kenji Ohba, Hiroyuki Iwaki, Hideyuki Matsunaga, Shingo Suzuki, Hiroko Misawa, Keiko Ishizuka, Yutaka Oki, Jaeduk Yoshimura Noh1 and Hirotoshi Nakamura Second Division, Department of Internal Medicine, Hamamatsu University School of Medicine, 1-20-1 Handayama, Hamamatsu, Shizuoka 431-3192, Japan 1Ito Hospital, 4-3-6 Jingumae, Shibuya-ku, Tokyo, 150-8308, Japan (Correspondence should be addressed to S Sasaki; Email: [email protected])

Abstract Transcriptional repression of the TSH-specific b subunit between ERa and GATA2 were mapped to the DNA-binding (TSHb) gene has been regarded to be specific to thyroid domain (DBD) of ERa and the Zn finger domain of GATA2. hormone (tri-iodothyronine, T3) and its receptors (TRs) in E2-dependent inhibition requires the ERa amino-terminal physiological conditions. However, TSHb mRNA levels in the domain but not the tertiary structure of the second Zn finger pituitary were reported to decrease in the administration of motif in E2-ERa-DBD. In the thyrotroph cell line, TaT1, E2 pharmacologic doses of estrogen (17-b-estradiol, E2)and treatment reduced TSHb mRNA levels measured by the reverse increase in E2 receptor (ER)-a null mice. Here, we investigated PCR. In the human study, despite similar free the molecular mechanism of inhibition of the TSHb gene thyroxine levels, the serum TSH level was small but significantly expression by E2-bound E2- 1 (E2-ERa). In higher in post- than premenopausal women who possessed no kidney-derived CV1 cells, transcriptional activity of the TSHb anti-thyroid antibodies (1.90 mU/mlG0.13 S.E.M.vs was stimulated by GATA2 and suppressed by THRBs 1.47 mU/mlG0.12 S.E.M., P!0.05). Our findings indicate and ERa in a ligand-dependent fashion. Overexpression of redundancy between T3-TR and E2-ERa signaling exists in PIT1 diminished the E2-ERa-induced inhibition, suggesting negative regulation of the TSHb gene. that PIT1 may protect GATA2 from E -ERa targeting by 2 Journal of Endocrinology (2008) 199, 113–125 forming a stable complex with GATA2. Interacting surfaces

Introduction Boado et al. (1983) showed that E2 benzoate induced a marked depression of intrapituitary TSH. Sekulic et al. (1998) Thyrotropin (thyroid-stimulating hormone, TSH) is a demonstrated that E2 treatment reduced the number of heterodimer consisting of the TSH-specific b subunit TSHb-positive cells in the rat pituitary. (TSHb) and the chorionic gonadotropin a chain (CGA) There are two types of estrogen receptors (ERs), ERa and that is common to luteinizing hormone (LH), follicle- ERb. In the anterior lobe of the pituitary, ERa is the major stimulating hormone (FSH), and chorionic gonadotropin. receptor (Kuiper et al.1996, Pelletier et al.2000, Liu & Cui Thyroid hormone (tri-iodothyronine, T3)-mediated negative 2005) and is expressed in thyrotrophs as well as gonadotrophs feedback of TSH production in the pituitary is a central (Stefaneanu et al.1994, Gittoes et al.1997). Interestingly, Scully mechanism of the pituitary–thyroid axis and is believed to be et al.(1997)reported that, in the pituitary of ERa-deficient specific to T3 at physiological concentrations. However, mice, mRNA levels for not only Lhb and Fshb but also Tshb and administration of pharmacologic doses of estrogen (17-b- Cga were dramatically elevated. Compared with wild-type estradiol, E2) was known to augment the effect of thyroid mice, the amount of mRNA for Tshb and Cga in Esra null mice hormone replacement to suppress TSHb and CGA mRNA increased by 3.20- and 4.36-fold respectively. Immunostaining (Wondisford 1996 and references therein). Indeed, E2 inhibits using specific antibodies revealed that the numberof TSHb-and the upregulation of Tshb and Cga mRNA in the pituitary of CGA-positive cells also increased (Scully et al.1997). These hypothyroid rats (Franklyn et al. 1987). Bottner & Wuttke findings are similar to observations in mice devoid of all known (2005) and Bottner et al. (2006) reported that ovariectomy T3 receptors (TRs; Gothe et al. 1999). Although these increased pituitary TSHb mRNA levels and that this observations suggest that E2 suppresses TSHb expression via elevation was abolished by 17b-E2-3-benzoate treatment. ERa, the molecular mechanism has not been clarified.

Journal of Endocrinology (2008) 199, 113–125 DOI: 10.1677/JOE-08-0128 0022–0795/08/0199–113 q 2008 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access 114 K NAGAYAMA and others . E2 inhibits TSHb

ER and TR belong to the nuclear GATA2 in thyrotrophs is essential for the expression of the superfamily and share a basic structure consisting of a TSHb gene (Gordon et al. 1997, Dasen et al. 1999). PIT1 is a receptor-specific amino-terminal domain (NTD), central pituitary-specific expressed in somato- DNA-binding domain (DBD), and carboxyl (C)-terminal trophs, lactotrophs, and thyrotrophs. GATA2 is a subtype of ligand-binding domain (LBD; Mangelsdorf et al. 1995). In the the GATA family of transcription factors and binds with the promoter region of target whose transcription is GATA-responsive element (GATA-RE) through its Zn finger enhanced by treatment with E2 or T3, ER homodimers and domain, which has high homology among all GATA family TR heterodimers formed with the members (Ferreira et al. 2005). It was recently reported that (RXR) recognize the E2-response element (ERE) and the TRAP220, which is a for liganded TR and ER, T3-response element (TRE) respectively. The tertiary also functions as a coactivator for GATA2 and PIT1 (Gordon structure of DBDs is maintained by two Zn finger motifs, et al. 2006) and plays an important role in the transactivation and the sequences of the P- and D-boxes of DBD of the TSHb gene (Ito et al. 2000). In addition, TRAP220 play critical roles in the differential recognition of ERE and was reported to be a coactivator for GATA1 (Stumpf et al. TRE (Umesono & Evans 1989). E2-bound ER (E2-ER) and 2006), transactivation by which is known to be inhibited by T3-bound TR (T3-TR) recruit coactivators such as members E2-ERa (Blobel et al. 1995, Blobel & Orkin 1996). of the p160 family or CREB-binding (CREB)/p300, Using non-pituitary CV1 cells, we recently reported that whose histone acetyl-transferase activities relax the chromatin T3-TRb2 represses the expression of the TSHb gene driven by structure and result in the enhancement of transcription GATA2 and PIT1 through interference of GATA2-induced (Perissi & Rosenfeld 2005). In addition, T3-TR (Ito & transactivation, and that, contrary to the previous reports, the Roeder 2001) and E2-ER (Zhang et al.2005) recruit negative TRE (nTRE; Wondisford et al. 1989) and negative TRAP220, a component of the TRAP/SRB/MED- regulatory elements (NRE; Sasaki et al. 1999) were not containing cofactor complex of the RNA polymerase II required (Matsushita et al. 2007). The reporter assay using CV1 holo-enzyme. In the absence of T3, TR recruits co-repressors cells provides an ideal experimental platform to investigate the such as co-repressor or the silencing difference between positive and negative regulation by T3, of retinoic acid and thyroid hormone receptors since this cell line is one of those most frequently used for the (SMRT). These co-repressors associate with histone deace- studyof positive regulation (Nakano et al. 2004). Wefound that tylases (HDACs), resulting in repression of transcription THRB-DBD directly interacts with the Zn finger domain of (Perissi & Rosenfeld 2005). GATA2 (GATA2-Zf) and that T3-TR targets TRAP220, Of note, TR and ER have the potential to bind an identical which functions as a coactivator for GATA2 in the context of half-site sequence, AGGTCA, although the number of the TSHb promoter (Matsushita et al. 2007). Reflecting T3- spacing between the half-sites and their orien- specificity in vivo, ligand-bound , vitamin tations are different between TRE and ERE (Mangelsdorf D receptor, RXR, or peroxisome proliferator-activated et al. 1995); ERE can be recognized by the TR monomer, in receptor g2 all did not exhibit negative regulation of the addition to homodimer and heterodimer with RXR (Klinge TSHb promoter in our experimental system (Nakano et al. et al. 1997). T3-TR can also directly bind to ERE of the 2004). Unexpectedly, however, E2-ERa exhibited significant promoter for the receptor gene and stimulate its suppression of the TSHb promoter although the magnitude of transcription (Scott et al. 1997). On the other hand, when TR suppression (39.4%G4.5 S.E.M.) was less than that of T3-TRb2 and ER coexist, T3-TR was reported to inhibit E2-ERa- (64.1%G2.6 S.E.M.; Nakano et al. 2004). Here, we report that, mediated transactivation of the promoters for the preproen- in the negative regulation of the TSHb gene, there is a kephalin gene (Zhu et al. 1996, Vasudevan et al. 2001b) and redundancy between E2-ERa- and T3-TR-mediated signal- the prolactin gene (Pernasetti et al. 1997), probably via ing pathways. squelching of coactivators common to both receptors. Similar mutual inhibition has been reported between T -TR and 3 Materials and Methods other ligand-bound hormone receptors (Zhang et al. 1996). Thus, the redundancy of DNA recognition and the Plasmid construction common utilization of cofactors have been postulated to mediate crosstalk between the E2-ER and T3-TR signaling We recently found that not only T3-TR (Tillman et al. 1993, pathways (Vasudevan et al. 2001a, 2002). Maia et al. 1996) but also E2-ERa have a tendency to suppress In the mouse TSHb promoter, the DNA sequence firefly luciferase-based reporter gene (data not shown). Thus, between nt K271/K80 (corresponding to the sequence we employed the CAT-based reporter system. The TSHb- between nt K269/K78 in the human TSHb gene) was CAT reporter gene was constructed by fusing the human reported to be sufficient for maximal promoter activity in TSHb promoter (nt. K128/C37) with the CAT reporter thyrotrophs (Wood et al. 1990). In this promoter region, there gene, whose backbone lacks the pUC-derived AP-1 site are binding sites for two transcription factors, PIT1 and (Sasaki et al. 1999, Nakano et al. 2004). TSHb-D4-CAT was GATA2 (Haugen et al. 1996, Gordon et al. 1997, 2002, Dasen reported previously (Matsushita et al. 2007). The DNA et al. 1999, Charles et al. 2006). Co-existence of PIT1 and fragment encompassing the promoter region for TSHb gene

Journal of Endocrinology (2008) 199, 113–125 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access E2 inhibits TSHb gene expression . K NAGAYAMA and others 115

(nt. 615/129) was amplified using PCR from human genomic GST pull-down assay DNA as template and subcloned into EcoRI site in TSHb- Escherichia coli (DH5a)transformedwithpGEX-4T-1- CAT to generate TSHb (615/C37)-CAT.InTSHb-M1- GATA2-Zf were induced with 0.1 mM isopropyl-1-thio-b- CAT, the putative nTRE/NRE remained intact (data not galactopyranoside for 4 h. The E. coli pellet was sonicated and shown). The expression plasmids for mouse Gata2 (pcDNA3- C the fusion were mixed with glutathione-sepharose mGATA2), human PIT1 (pCB6 -hPit1), wild-type human beads (Amersham Pharmacia Biotech) for purification. TRb1 (pCMX-hTRb1), TRb1- mutants (pCI-C1 Receptor proteins were translated in vitro using rabbit and C2), wild-type rat Trb2 (pCMX-rTRb2), and human reticulocyte lysate (Promega Corp.) in the presence of ERa were described previously (Nakano et al. 2004). The 35S-. Radio-labeled receptors were incubated deletion mutants of human ERa (D170 and D247) and mutant with GST protein fused to GATA2-Zf (GST-Zf) in the ERa such as C205S, C221G, and C240S were generated from binding buffer (150 mM NaCl, 20 mM Tris–HCl (pH 7.5), a wild-type ER using standard subcloning techniques or a site- 0.3% Nonidet P-40, 1 mM dithiothreitol, 0.5 mM phenyl- directed mutagenesis kit (Stratagene). The plasmid expressing methylsulfonyl fluoride, 2 mg/ml aprotinin and leupeptin) for the recombinant protein of the GATA2 Zn finger domain 3 h at 4 8C. The precipitants were washed three times with (GATA2-Zf) combined with glutathione-S-transferase the binding buffer. Bound protein was analyzed by 10% SDS- (GST), pGEX-4T-1-GATA2-Zf, was described elsewhere PAGE and visualized using the FLA-3000 autoradiography (Matsushita et al. 2007). All constructs were confirmed by system (Fujifilm, Tokyo, Japan). DNA sequencing.

Gel shift assay Cell culture and transient transfection Oligonucleotides for vit-A2 consensus ERE (Scott et al. 1997; CV1 cells were grown in monolayers and cultured at 37 8C sense, 50-AATTCGTCCAAAGTCAGGTCACAGTGAC- under CO2/air (1:19) in Dulbecco’s modified Eagle’s medium CTGATCAAAGTT-30; antisense, 50-AACTTTGATCAG- (DMEM) containing 10% dextran–charcoal-stripped fetal GTCACTGTGACCTGACTTTGGACGAATT-30)were bovine serum (10% DCC serum), penicillin G (100 units/ml), 32 labeled with g P-ATP using thyroxine (T4)-polynucleotide and streptomycin (100 mg/ml). TaT1 cells, a mouse thyro- kinase (Toyobo, Tokyo, Japan). The receptor proteins were troph cell line (Yusta et al. 1998), were seeded on Matrigel- translated in vitro using a TNT T7 quick-coupled transcrip- coated plates (Becton Dickinson Labware, Bedford, MA, tion/ system (Promega Corp). The radio-labeled USA). The cells were maintained under the same conditions as probes and in vitro-translated receptors were incubated for CV1 cells. The CV1 cells were trypsinized and plated in 20 min at room temperature in 20 ml binding buffer containing 60 mm diameter dishes for 24 h before transient transfection 10 mM HEPES–NaOH (pH 7.9), 50 mM KCl, 0.1mM using the calcium phosphate technique (Sasaki et al. 1995). EDTA, 10% glycerol, 0.5 mM dithiothreitol, and 0.1 mg/ml 6 The cells at a density of 10 cells/plate were transfected with poly (dI-dC). The DNA–protein complexes were resolved on . 0 6 mg of the expression plasmids for Esra or Thrb along with 5% polyacrylamide gels at 150 V for 1.5 h. Gels were dried and 4.0 mg of the Tshb-Cat reporter gene, 1.8 mg b-galactosidase visualized using the FLA-3000 autoradiography system. expression vector pCH111 (a modified version of pCH110; Pharmacia LKB Biotechnology), 0.1 mg human PIT1 C expression vector (pCB6 -hPit1), and 0.4 mg mouse Gata2 RNA isolation and real-time reverse transcription-PCR expression vector (pCDNA3-mGATA2), and the pCMX (RT-PCR) empty vector as carrier DNA (7.2 mg DNA per dish in total). TaT1 cells cultured in 10% DCC serum were incubated in the After the cells were exposed to the calcium phosphate/DNA presence of 1 mME2 or T3 for 48 h. The cells were harvested precipitates for 20 h, the medium was replaced with phenol and total RNAs were purified by the acid guanidium red-free DMEM containing 5% (v/v) and the cells were thiocyanate–phenol–chloroform extraction method (Kawai allowed to grow for an additional 24 h with or without 1 mM et al. 2004). For the first-strand cDNA synthesis, 2 mg total E2 or T3. After incubation for an additional 24 h, the cells were RNAwere mixed with random hexanucleotides and 200 units harvested and CAT activity was measured as described Moloney murine virus reverse transcriptase (Invi- previously (Sasaki et al. 1995). The transfection efficiency trogen Corp). The cDNA for TSHb was amplified with the was normalized by b-galactosidase assay. For each CAT forward primer (50-GGCAAACTGTTTCTTCCCAA-30) reporter assay, we performed transfection with CAT reporter and the reverse primer (50-TCTGTGGCTTGGTGCAG- gene driven by cytomegalovirus promoter (CMV-CAT; TAG-3 0). The cDNA for glyceraldehyde-3-phosphate 40 ng/dish), the magnitude of which was adjusted to the dehydrogenase (GAPDH) was amplified with the forward value of 100. Immunoblotting with anti-ERa antibody primer (50-TGAACGGGAAGCTCACTGG-30)andthe (1:200, cat. sc-8002, Santa Cruz Biotechnology, Santa Cruz, reverse primer (50-TCCACCACCCTGTTGGCTGTA-30). CA, USA) was performed using a standard method described PCR amplification was carried out using a DNA thermal previously (Matsushita et al. 2007). cycler (Takara Bio Inc., Shiga, Japan) under the following www.endocrinology-journals.org Journal of Endocrinology (2008) 199, 113–125

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access 116 K NAGAYAMA and others . E2 inhibits TSHb gene expression

conditions: denaturation at 95 8C for 1 min, annealing at 62 8C (Gordon et al. 1997, Dasen et al. 1999), co-expression of PIT1 for 1 min, extension at 72 8C for 2 min, and the final extension and GATA2 transactivated the TSHb promoter in CV1 cells at 72 8C for 4 min. We determined the cycle number for (Fig. 1B, lane 2). This activity was significantly repressed by each primer set so that the specific product was amplified not only T3-TRb2 but also E2-ERa (lanes 2, 4, and 6). during the exponential phase of the amplification. Based on When cognate receptors were co-expressed with PIT1 and preliminary studies (data not shown), 27 cycles were employed GATA2, T3 and E2 inhibited transcription by 64.1%G2.6 for the amplification of TSHb and GAPDH cDNA. The PCR S.E.M. (lanes 3 and 4) and 39.4%G4.5 S.E.M. (lanes 5 and 6) products were subjected to electrophoresis on a 1.4% agarose respectively. E2-ERa-induced repression depended on the gel and stained with ethidium bromide. Using the SYBR E2 concentration (Fig. 1C) and the expression level of ERa Green I kit and Light Cycler (Roche Diagnostics), precipitated (Fig. 1D). The inhibition of the TSHb promoter by E2-ERa DNA was quantified by real-time PCR using the primers was observed in Hela and 293T cells (data not shown). We mentioned above. The thermal cycling conditions were also examined the effect of E2-ERa on the TSHb 10 min at 95 8C, followed by 27 cycles of 10 s at 95 8C for (K615/C37)-CAT reporter gene, which has longer TSHb denaturing, 10 s at 62 8C for annealing, and 7 s at 72 8C for promoter encompassing nt. K615/C37, and found again extension. PCR signals were analyzed using Light Cycler that the transcriptional activity in CV1 cells was stimulated by software Ver. 3.5 (Roche Diagnostics). PIT1 and GATA2, and significantly decreased by 1 mME2 (Fig. 1E).

Measurement of serum TSH, free T4, and E2 in pre- and postmenopausal women Overexpression of PIT1 relieves E2-dependent inhibition of the TSHb gene and E -ERa targets GATA2-induced Among female patients who visited Ito Hospital in May 2004, 2 transactivation 134 subjects with simple goiter or thyroid nodule(s) and no thyroid autoantibodies against thyroglobulin, thyroid per- We tested the effect of PIT1 overexpression on the negative oxidase, or TSH receptor were extracted. They were divided regulation of the TSHb promoter by E2-ERa. As shown in into 62 premenopausal women and 72 postmenopausal Fig. 2A, inhibition by E2-ERa was abolished by increased women. The premenopausal group consisted of women levels of PIT1, suggesting that a large amount of PIT1 between 30 and 50 years old (37.35G5.40 years old) and their antagonizes the E2-ERa-dependent inhibition of the TSHb serum E2 levels were all higher than 20 pg/ml. The gene. We recently found that deletion of a short sequence postmenopausal group included women between 50 and 70 between GATA-REs and the TATA-box in the TSHb (59.29G4.70) years old with serum E2 levels lower than promoter (nt. 82/52; Fig. 1A) enabled GATA2 alone to 20 pg/ml. TSH and free T4 were measured with Roche transactivate without PIT1 (Matsushita et al. 2007), and we ECLusys kit (Roche) and E2 with electro chemiluminescence designated this deleted sequence as the suppressor region immunoassay (SRL, Tachikawa, Japan). (SR). The construct TSHb-M1-CAT (Fig. 2B, left panel) in which SR was deleted was activated by GATA2 alone, and this activity was inhibited by E -ERa (Fig. 2B, right panel). Statistical analysis 2 GATA2 alone without PIT1 transactivated two other reporter Each experiment was performed in duplicate more than three constructs, TSHb-D4-CAT, in which nTRE/NRE in different times and each result is expressed as the meanGS.E.M. TSHb-M1-CAT was mutated (Matsushita et al. 2007), and Using StatView 4.0 software (Abacus Concepts Inc., CGA-CAT, which possesses a functional GATA-RE (Steger Berkeley, CA, USA), we examined statistical significance in et al. 1994). Again, E2-ERa repressed the GATA2-induced Figs 6 and 7 by Mann–Whitney U test. All other statistical transactivation of TSHb-D4-CAT and CGA-CAT by analyses were performed using one-way ANOVA followed by 53.1%G1.9 S.E.M. and 30.5%G6.2 S.E.M. respectively (data Fisher’s protected least significant difference test. P!0.05 was not shown). Collectively, transactivation by GATA2 alone was considered significant. sufficient to mediate negative regulation by E2-ERa and PIT1 antagonized this inhibition.

Results The Zn finger domain of GATA2 directly interacts with TR and ERa in vitro TSHb promoter activity is inhibited by not only T -TRb2 but 3 The amino acid sequence of the Zn finger domain of also E -ERa 2 GATA2 (GATA2-Zf), but not the N- or C-terminal region, To explore the molecular mechanism underlying the E2- has high homology with that of GATA1 (Fig. 3A), and both dependent inhibition of the TSHb gene expression, we GATAs are known to transactivate the same GATA-RE. In carried out a reporter assay using TSHb-CAT (Fig. 1A), in addition, E2-ERa was reported to inhibit GATA1-induced which the human TSHb promoter (nt. 128/C37) is fused to transactivation (Blobel et al. 1995). As predicted, GATA1- the CAT reporter gene. In accordance with previous reports induced transactivation of TSHb-CAT was also repressed by

Journal of Endocrinology (2008) 199, 113–125 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access E2 inhibits TSHb gene expression . K NAGAYAMA and others 117

Figure 1 Effect of E2-ERa on the TSHb promoter activity stimulated by PIT1 and GATA2. (A) Schematic of the TSHb-CAT. The binding sites for PIT1 and GATA2, suppressor region (SR), TATA box, and reported nTRE/NRE are indicated as boxes. (B) Using the calcium phosphate method, the expression plasmids for ERa or TRb2(0.6 mg) were co-transfected with TSHb-CAT (4.0 mg), PIT1 (0.1 mg), and GATA2 (0.4 mg) into CV1 cells in the absence or presence of 1 mME2 or T3 respectively. After incubation for 24 h, the cells were harvested and the CAT activity was measured. The CAT activity of CMV-CAT is represented as 100%. The data are shown as the meanG # S.E.M. of at least six individual experiments. *P!0.05; **P!0.01; P!0.05 versus PIT1 and GATA2. (C) TSHb-CAT was co-transfected ! with PIT1, GATA2, and ERa into CV1 cells in the presence of 0–1 mME2. The CATactivity of CMV-CATis represented as 100% *P 0.05; **P!0.01. (D) Increasing amounts of the ERa expression plasmid (0–0.6 mg/6 cm dish) were co-transfected with TSHb-CAT (4.0 mg) and the expression plasmid for PIT1 (0.1 mg) and GATA2 (0.4 mg) into CV1 cells. The magnitude of CAT activity without E2 was divided by G that with E2 (1 mM) to calculate the fold repression. The results are shown as the means S.E.M. from six independent experiments. *P!0.05; **P!0.01. (E) The expression plasmids for ERa (0.6 mg) were co-transfected with TSHb (K615/C37)-CAT (4.0 mg), PIT1 G (0.1 mg), and GATA2 (0.4 mg) into CV1 cells in the absence or presence of 1 mME2. All results in (E) are shown as means S.E.M. for eight independent experiments. *P!0.05. The CAT activity of CMV-CAT is represented as 100%.

E2-ERa (Fig. 3B). While C-terminal finger of GATA1 induced by GATA2-C295A was not inhibited by E2-ERa. directly recognizes the GATA-RE, the N-terminal finger is These findings indicate that interaction between GATA2-Zf known to interact with Friend of GATA (FOG) 1 and 2 that and ERa is essential to E2-ERa-mediated inhibition. Using are the strong co-repressor for the GATA family transcrip- the GST fusion protein with GATA2-Zf (GST-Zf, Fig. 3A), tion factors. We reported (Matsushita et al. 2007) that we tested the direct interaction between ERa and GATA2- transactivation by a mutant GATA2, GATA2-C295A Zf in vitro. As reported previously (Matsushita et al. 2007), (Fig. 3A), which has an amino acid substitution from GST-Zf interacted with TRb1(Fig. 3D, upper panel) but cysteine to alanine at codon 295 of the N-terminal Zn not the luciferase protein (lower panel). In an E2- finger, was resistant to the inhibition by T3-TRb2. The independent manner, GST-Zf also bound to radio-labeled finding that the basal transcription by C295A was full-length ERa (Fig. 3D, middle panel). As shown in significantly higher than that by empty vector (Fig. 3C) Fig. 3E, the mutant ERas N2 and D247, which lacked indicates that the overall structure as functional GATA2 is DBD, did not interact with GST-Zf, suggesting that ERa- partly preserved in this mutant. However, transactivation DBD binds to GATA2-Zf as in the case of THRB-DBD. www.endocrinology-journals.org Journal of Endocrinology (2008) 199, 113–125

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access 118 K NAGAYAMA and others . E2 inhibits TSHb gene expression

The E2-induced inhibition of the TSHb promoter requires ERa-NTD We previously reported that TRb1-NTD is dispensable for the negative regulation of the TSHb promoter (Nakano et al. 2004). We wanted to compare the function of NTDs of ERa. In contrast to TRb1, the ERa-NTD truncation mutant D170 (Fig. 5A, middle upper panel) did not inhibit but rather stimulated the transcription (right panel), although it mediated positive regulation (left panel). These data suggest that ERa-NTD is critical for inhibition by E2. As predicted, ERa (D247), which lacks the DBD, did not exhibit ligand- dependent transactivation or inhibition. The expression levels of mutant ERas were comparable (middle lower panel). As shown in Fig. 5B, the amino acid sequences between ERa and ERb were conserved in their DBD (96%) and LBD (58%) but not in NTDs (McInerney et al. 1998). We found Figure 2 Roles of PIT1 in the negative regulation of TSHb gene that repression of the TSHb promoter by E2-ERb was expression by E2-ERa (A) TSHb-CAT (4.0 mg) was transfected into CV1 cells together with the expression plasmids for PIT1 (0.1– impaired and the effect of E2 was not statistically significant 0.4 mg), GATA2 (0.4 mg), and ERa (0.6 mg) in the absence or (Fig. 5C). To exclude the possibility that ERa-LBD may presence of 1 mME2 (left panel). The fold repression (right panel) communicate with the NTD of the same ERa molecule, we was calculated from CAT activity without E2 divided by that with tested the function of the chimeric ERa, bNaC, in which 1 mME2. (B) Schematic of TSHb-M1-CAT (left panel). The suppressor region (SR) was deleted in this construct. TSHb-M1-CAT the NTD was substituted with that of ERa (Yi et al. 2002; was co-transfected with the expression plasmids for ERa,GATA2 Fig. 5B). Again, bNaC failed to mediate E2-dependent into CV1 cells in the absence or presence of 1 mME2. The results are repression (Fig. 5D). Western blotting with an antibody G ! . shown as means S.E.M. from six experiments. *P 0 05. against the C-terminal region of ERa indicated that expression levels of wild-type ERa and bNaCwere comparable (Fig. 5E). Together, these findings indicate that The E2-ERa-induced inhibition of the TSHb promoter does not require the second Zn finger motif the ERa-NTD has an important role in E2-dependent inhibition. P- and D-boxes and the zinc finger structure in DBD are critical constituents that determine receptor specificity in ligand- dependent positive regulation (Mangelsdorf et al.1995). To Effect of E2 on the expression of the TSHb gene in thyrotroph cell investigate the involvement of these motifs in the negative line TaT1 b a regulation of the TSH gene, we generated three mutant ER s: Serum E2 concentration is known to influence not only the C205S in the P-box and C221G and C240S in the stem of the level of T4-binding globulin (TBG; Surks & Sievert 1995)but second Zn finger (Fig. 4A, upper panel). Western blot analysis also the production of T3 and T4 from the thyroid gland with an anti-ERa antibody indicated that expression levels of (Furlanetto et al. 1999, Sosic-Jurjevic et al. 2005, Alotaibi et al. these mutants in CV1 cells were comparable (lower panel). As 2006, Lima et al. 2006). Using cultured thyrotroph cell line predicted, these mutants lost DNA-binding affinity as demon- TaT1 (Yusta et al. 1998), we studied the direct effect of E2 on strated by gel shift assay with vitellogenin A2-derived ERE (vit- the TSHb mRNA level. When TSHb mRNA was detected A2-ERE; Scott et al.1997; Fig. 4B), and failed to activate the by conventional RT-PCR, not only T3 but also E2 reduced ERE-tk-CAT reporter gene, in which vit-A2 ERE is fused to the intensity of the TSHb mRNA band (Fig. 6A). We further the CAT reporter gene driven by the thymidine kinase (tk) measured TSHb mRNA expression by real-time quantative promoter (Fig. 4C, left panel). As shown in the right panel of RT-PCR(Fig. 6B). Although the magnitude was smaller than Fig. 4C, negative regulation of the TSHb promoter induced by that for T3,E2 treatment also significantly reduced the level of PIT1 and GATA2 was abolished in C205S and C221G. The TSHb mRNA (67.2%G12.1 S.E.M.vs49.5%G14.9 S.E.M., findings that C205S and C221G directly interacted with P!0.05). The magnitudes of suppression were correlated to GATA2-Zf (Fig. 4D) suggest that physical interaction of those from the reporter assay using TSHb-CAT (Fig. 1B). ERa-DBD with GATA2 is not sufficient for inhibition by E -ERa. Unexpectedly,another mutant ERa,C240S(Fig. 4A) 2 Comparison of serum TSH levels between pre- and post- preserved E -induced negative regulation of the TSHb gene 2 menopausal women (Fig. 4C) as well as in vitro binding with GATA2-Zf (Fig. 4D), suggesting that the tertiary structure of the second Zn finger To evaluate the clinical significance of E2 effects on serum motif, which is necessary for ERE recognition, is dispensable TSH, we compared the serum TSH level of 72 postmeno- for inhibition of the TSHb gene. pausal women with that of 62 premenopausal women,

Journal of Endocrinology (2008) 199, 113–125 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access E2 inhibits TSHb gene expression . K NAGAYAMA and others 119

Figure 3 The interaction between GATA2-Zf and ERa-DBD. (A) Schematic of GATA1, GATA2, GATA2-C295A, and GST-Zf. The boundaries of the Zn finger region were deduced by comparison with chicken GATA1 reported by Boyes et al. (1998). The homology of amino acids among GATA1 relative to GATA2 is indicated as a percentage in the boxes. In GST-Zf, GST protein was fused to GATA2-Zf. (B) The transactivation of TSHb-CAT by GATA1 and GATA2 was suppressed by E2-ERa. The expression plasmids for mouse GATA1 or 2 (0.4 mg) were transfected together with that for PIT1 (0.1 mg) into CV1 cells. The results are shown as the meansGS.E.M. from six independent experiments. (C) The expression plasmids for mouse GATA2 or GATA2-C295A (0.4 mg) were transfected into CV1 cells under the same conditions as Fig. 1B. The results are shown as the meansGS.E.M. from six independent experiments. The magnitude of the CAT activity stimulated by wild-type GATA2 and PIT1 was taken as 100%. *P!0.05; **P!0.05 versus reporter alone (first lane). (D) GST-Zf was incubated with 35S-labeled TRb1orERa in the presence or 35 35 absence of 1 mMT3 or E2 respectively. S-labeled luciferase protein was utilized as a control. Arrowheads indicate S-labeled TRb1 (upper panel), ERa (middle panel), and luciferase protein (lower panel). luc: luciferase protein. I and B indicate input and bound respectively. (E) 35S-labeled wild-type ERa or its deletion constructs were incubated with GST-Zf. GST-Zf interacted with full-length ERa, D170, and N1 but not with D247 or N2, which lack DBD. I and B indicate input and bound respectively. who visited Ito Hospital because of simple goiter or Discussion thyroid nodule(s) in May 2004 and none had any thyroid autoantibodies. Because age-related decline in thyroid In the present study, we demonstrated that E2-ERa inhibits function is not noticeable until the eighth decade of life transactivation of the TSHb gene through a mechanism (Mariotti et al. 1993), we recruited female subjects with ages similar to inhibition by T3-TRs. We used reporter assay in the up to 70 years old as the postmenopausal population. As CV1 cell line, which is ideal for comparing the mechanism of predicted, the serum E2 levels in postmenopausal females ligand-dependent negative regulation of the TSHb gene with were very low (Fig. 7, left panel). Although there was no that of positive regulation (Nakano et al. 2004). Our findings difference in the free T4 concentrations (middle panel), the imply that redundancy in the recognition of target genes exists serum TSH level of the postmenopausal group was not only in the positive regulation (Klinge et al. 1997, Scott significantly higher than that of the premenopausal group et al. 1997) but also in the negative regulation of the TSHb (right panel). gene. It was postulated that E2-dependent inhibition of the www.endocrinology-journals.org Journal of Endocrinology (2008) 199, 113–125

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access 120 K NAGAYAMA and others . E2 inhibits TSHb gene expression

Figure 4 analysis of ERa-DBD. (A) Schematic of ERa-DBD (upper panel). Cysteines at codon 205 in the P-box and 221 and 240 in the stem of second Zn finger were substituted to (C205S), (C221G), and serine (C240S). The expression levels of wild-type and mutant ERas were comparable (lower panel). CV1 cells were transfected with the expression plasmids for wild-type or mutant ERas. Whole cell extracts were fractionated by SDS-PAGE and subjected to western blot (WB) with anti-ERa antibody against the C-terminal region of ERa. (B) 32P-radio-labeled vit-A2 ERE was incubated with in vitro-translated protein for wild-type or mutant ERas. ERE, specific cold competitor. NS, non-specific cold oligo DNA. Arrow, specific binding of ERa; arrowhead, nonspecific binding. (C) ERE-tk-CAT (left panel) was transfected into CV1 cells together with the expression plasmid for ERa in the presence or absence of 1 mME2. CAT activity with 1 mME2 was divided by that without E2 to calculate fold activation. TSHb-CAT (right panel) was transfected into CV1 cells together with the expression plasmids for PIT1, GATA2, and wild or mutant 35 ERas as shown in Fig. 1B. CATactivity without E2 was divided by that with 1 mME2 to calculate fold repression. (D) S-radio-labeled receptors were incubated with GST protein fused to GST-Zf and binding fractions were subjected to SDS-PAGE.

TSHb gene was mediated via direct binding of ERa with (Chuang et al. 1997, Schaufele 1999), whose most important the putative nTRE (Wondisford 1996). However, this is PIT1-binding site is located adjacent to the ERE unlikely, because E2-induced inhibition is preserved in the (Nowakowski & Maurer 1994). However, current results ERa mutant C240S, which fails to bind or activate indicate that E2-ERa-dependent inhibition of the TSHb canonical vit-A2-ERE (Fig. 4C). In agreement with this, promoter does not require PIT1 (Fig. 2B) and is rather crippled E2-ERa inhibited GATA2-induced transactivation of by the overexpression of this transcription factor (Fig. 2A). TSHb-D4-CAT, which lacked a DNA sequence homologous Although the results of the transfection experiments as well to the half-site. Moreover, E2-ERa repressed the activity of as the cell culture studies are obtained with high pharma- TSHb-M1-CAT (Fig. 2B) and CGA-CAT stimulated by cological doses of E2, the overall results were in accordance GATA2 alone (data not shown), suggesting that interference with our recent findings regarding the T3-dependent of GATA2-induced transactivation by E2-ERa is the inhibition of the TSHb gene (Matsushita et al. 2007). The mechanism for inhibition of the TSHb gene. This notion is repression of GATA2-induced activation by E2-ERa and T3- further supported by the observation that transactivation by TR may occur by a ‘tethering mechanism’ (Nissen & the mutant GATA2-C295A was resistant to E2-ERa-induced Yamamoto 2000, Herrlich 2001), where liganded nuclear suppression (Fig. 3C). ERa interacts directly with PIT1 receptors interfere with the activity of DNA-binding (Ying et al. 1999), and these two transcription factors synergize transcription factors through protein–protein interactions. in the transactivation of the prolactin gene in lactotrophs In this kind of inhibition, redundancy among liganded

Journal of Endocrinology (2008) 199, 113–125 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access E2 inhibits TSHb gene expression . K NAGAYAMA and others 121

Figure 5 Involvement of NTD in the negative regulation of TSHb gene by E2-ERa. (A) In the presence of 1 mME2, expression plasmids for ERa or its mutants, D170 and D247 (middle upper panel) were transfected into CV1 cells together with the ERE-tk-CAT reporter gene (left panel) or TSHb-CAT and the expression plasmids for PIT1 and GATA2 (right panel). *P!0.05 versus empty vector for receptor expression plasmids. The CAT activity with the ligand was divided by that without ligand to calculate fold activation (left panels). The CAT activity without ligand was divided by that with 1 mM ligand to calculate fold repression (right panels). The expression levels of wild-type and mutant ERas were comparable (middle lower panel). The CV1 cells were transfected with the expression plasmid for wild-type or mutant ERaD170 and D247). Western blot (WB) was performed as Fig. 4A. (B) Schematic of ERa, ERb and chimeric receptor bNaC. In bNaC, NTD of ERa is replaced with that of ERb. The homology of amino acids among ERb and bNaC relative to ERa is indicated as a percentage in the boxes. (C) The expression plasmid for ERa or ERb was transfected into CV1 cells together with TSHb-CAT and the expression plasmids for PIT1 and GATA2. #P!0.05 versus PIT1 and GATA2. (D) In the presence or absence of 1 mME2, the expression plasmid for ERa or bNaC was transfected into CV1 cells together with TSHb-CATand the expression plasmids for PIT1 and GATA2. CAT activity without E2 was divided by that with E2 (1 mM) to calculate the fold repression (right panel). The results are shown as meansGS.E.M. for eight separate experiments. *P!0.05. (E) Expression levels of wild-type ERa and bNaC were comparable. The CV1 cells were transfected with the expression plasmid for wild-type or bNaC and western blot with antibody against the C-terminal region of ERa was performed as Fig. 4A.

receptors has been reported. For example, E2-ERa the GST pull-down assay (Fig. 3E) indicated that DBD is (Kalaitzidis & Gilmore 2005) and (GC)- essential for the interaction of ERa with GATA2-Zf as bound GC receptor (GR; McKay & Cidlowski 1999, De in the case of THRBs (Matsushita et al. 2007). Second, E2- Bosscher et al. 2003) repress transactivation by the proin- dependent inhibition does not require the tertiary structure of flammatory transcription factor NF-kB by interacting with it. the second Zn finger motif in ERa-DBD, since E2- Similarly, the transcriptional activity of AP-1 (typically the dependent inhibition was preserved in the mutant ERa Jun/Fos heterodimer) is impaired by T3-TR and GC-GR C240S (Fig. 4C), which contains a disrupted second Zn (De Bosscher et al. 2003). finger motif (Fig. 4A). Third, it should be noted that the Although the precise mechanism of how E2-ERa inhibits mutant ERas, C205S, and C221G, failed to inhibit E2- the function of GATA2 in transactivation of the TSHb gene is dependent inhibition of the TSHb gene (Fig. 4C), although uncertain, the present analysis provides several insights. First, both mutants can physically bind GATA2-Zf in vitro www.endocrinology-journals.org Journal of Endocrinology (2008) 199, 113–125

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access 122 K NAGAYAMA and others . E2 inhibits TSHb gene expression

TRAP220 (Crawford et al. 2002, Gordon et al. 2006), and transcriptional repressors such as HDAC3 (Ozawa et al. 2001) and FOG 1 and 2 (Ferreira et al. 2005). ERa and TR share common coactivators including the p160 family (SRC1, TIF2/GRIP1, and AIB1/ACTR), CBP/p300, p300/CBP- associating factor (PCAF), and TRAP220. As in the case of T3-THRBs (Matsushita et al. 2007), overexpression of SRC-1, CBP, or PCAF did not show reproducible effects on the negative regulation of the TSHb gene by E2-ERa (data not shown). TRAP220 is required for the expression of the TSHb gene (Ito et al. 2000) and functions as a coactivator for GATA2 (Gordon et al. 2006), GATA1 (Stumpf et al. 2006), TR (Ito & Roeder 2001), and ERa (Zhang et al. 2005). TR (Ito & Roeder 2001) and ERa (Zhang et al. 2005) recognize an extended amino acid sequence encompassing two LXXLL motifs in TRAP220 with a specificity distinct from that of coactivators in the p160 family (Acevedo & Kraus 2003, Coulthard et al. 2003). We reported that dnTRAP220 (Yuan Figure 6 Effect of E2 on the expression of TSHb mRNA in the cultured thyrotroph cell line, TaT1. (A) TSHb mRNA isolated from et al. 1998) relieved the T3-TR-dependent inhibition of the TaT1 cells treated with 1 mMT3 or 1 mME2 for 48 h and evaluated TSHb promoter and that TRAP220 dissociated from this K with RT-PCR. RT ( ), PCR without reverse transcriptase. (B) TSHb promoter after T3 treatment of TaT1 cells (Matsushita et al. mRNA expression evaluated by real-time quantitative RT-PCR. The 2007). We found that the expression of the dnTRAP220 also experiments were repeated five times. The amount of the PCR a products was normalized with that of the GAPDH gene. The data abolished the E2-ER -induced inhibition of the TSHb gene are shown as meansGS.E.M. Statistical significance was determined (data not shown). Thus, TRAP220 may play a role in the ! by Mann–Whitney U test. *P 0.05. inhibition of the TSHb gene by E2-ERa and T3-TR. We found that ERa-NTD plays an essential role in the (Fig. 4D). The same results were obtained with the mutant repression of the TSHb gene by E2 (Fig. 5). This domain TRb2 G182E (Matsushita et al. 2007) and the mutant TRb1 interacts directly with the p160 family (Lavinsky et al. 1998, C127S and C145G (Nakano et al. 2004 and data not shown). Webb et al. 1998), p300 (Kobayashi et al. 2000), HDAC4 Physical binding alone is not sufficient to mediate ligand- (Leong et al. 2005), and repressor of tamoxifen transcriptional dependent inhibition, and subtle changes of amino acids on activity (Norris et al. 2002). Although ERa-NTD may bind the surface of DBD may be functionally critical. Similar directly with LBD (Kraus et al. 1995, Metivier et al. 2000, et al. conclusions were made after analysis of mutant GR using 2001) or regulate the interaction between the antagonist bound LBD and its co-repressor (Lavinsky et al. 1998), the circular dichroism (Tao et al. 2001). bNaC data indicate that ERa-NTD has an intrinsic function GATA2-Zf interacts with multiple coactivators including to control the E -dependent inhibition of the TSHb gene. CBP/p300 (Blobel et al. 1998, Hayakawa et al. 2004), 2 TRb2-NTD was reported to have a strong T3-independent transactivation effect on TRE (Sjoberg & Vennstrom 1995) and also to regulate the interaction between LBD and SMRT (Yang & Privalsky 2001). A function unique to TRb2-NTD may account for its stronger repression of the TSHb gene than TRb1(Nakano et al. 2004). Importance of GR-NTD was also reported for ligand-dependent inhibition of AP-1 activity (Heck et al. 1994). E2 is known to have profound influence on the thyroid hormone system, including cell proliferation (Furlanetto et al. 1999), iodine uptake (Furlanetto et al. 1999, Alotaibi et al. 2006, Lima et al. 2006), thyroid peroxidase activity (Lima et al. 2006), de-iodinases (Lisboa et al. 2001, Wasco Figure 7 Elevation of serum TSH levels after menopause in women et al. 2003), or TBG (Gross et al. 1971, Surks & Sievert 1995, with simple goiter who possess no autoantibodies against Arafah 2001). Altered thyroid hormone production and thyroglobulin, thyroid peroxidase, or TSH receptor. The levels of E 2 may mask the direct effect of E2 on TSHb (left panel), free T4 (middle panel), and TSH (right panel) in 62 expression. To exclude the influences of E on T and T premenopausal women and 72 postmenopausal women are 2 3 4 indicated. The data are shown as meansGS.E.M. Statistical levels in the circulating , we measured TSHb mRNA significance was determined by Mann–Whitney U test. *P!0.05; in TaT1 cells and found that E2 significantly repressed the **P!0.01. expression of TSHb mRNA (Fig. 6). We studied serum

Journal of Endocrinology (2008) 199, 113–125 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access E2 inhibits TSHb gene expression . K NAGAYAMA and others 123

TSH levels in women with simple goiter or thyroid Blobel GA & Orkin SH 1996 Estrogen-induced apoptosis by inhibition of the nodule(s), who had no thyroid autoantibodies. The TSH erythroid transcription factor GATA-1. Molecular and Cellular Biology 16 level was significantly higher in the postmenopausal than 1687–1694. Blobel GA, Sieff CA & Orkin SH 1995 Ligand-dependent repression of the premenopausal population (Fig. 7). Since the age of the erythroid transcription factor GATA-1 by the estrogen receptor. Molecular postmenopause group was between 50 and 70 years old, and Cellular Biology 15 3147–3153. thyroidal deterioration due to aging is unlikely (Mariotti Blobel GA, Nakajima T, Eckner R, Montminy M & Orkin SH 1998 CREB- et al. 1993). A small but significant increase in the TSH level binding protein cooperates with transcription factor GATA-1 and is after menopause may be related to the decreased TSH required for erythroid differentiation. PNAS 95 2061–2066. Boado R, Ulloa E & Zaninovich AA 1983 Effects of oestradiol benzoate on suppression by E2. Although free T4 or serum E2 data were the pituitary–thyroid axis of male and female rats. Acta Endocrinologica 102 not available, the NHANES III report (Hollowell et al. 2002) 386–391. also showed that the TSH concentration tended to increase De Bosscher K, Vanden Berghe W & Haegeman G 2003 The interplay with age in female populations without thyroid autoanti- between the and nuclear factor-kB or bodies. Surks & Hollowell (2007) have recently analyzed protein-1: molecular mechanisms for gene repression. Endocrine Reviews 24 TSH frequency distribution curves in NHANES III and 488–522. Bottner M & Wuttke W 2005 Chronic treatment with low doses of estradiol showed the progressive shift toward higher concentrations of affects pituitary and thyroid function in young and middle-aged TSH with age irrespective of thyroid antibodies. They say ovariectomized rats. Biogerontology 6 261–269. that an explanation for this shift to higher TSH ranges in Bottner M, Christoffel J, Rimoldi G & Wuttke W 2006 Effects of long-term elderly people is not apparent, but we consider that our treatment with resveratrol and subcutaneous and oral estradiol adminis- finding in this study is one of the main factors. tration on the pituitary–thyroid-axis. Experimental and Clinical Endocrinology and Diabetes 114 82–90. Boyes J, Byfield P, Nakatani Y & Ogryzko V 1998 Regulation of activity of the transcription factor GATA-1 by acetylation. Nature 396 594–598. Declaration of Interest Charles MA, Saunders TL, Wood WM, Owens K, Parlow AF, Camper SA, Ridgway EC & Gordon DF 2006 Pituitary-specific Gata2 knockout: effects on gonadotrope and thyrotrope function. Molecular Endocrinology 20 The authors declare that there is no conflict of interest that could be perceived 1366–1377. as prejudicing the impartiality of the research reported. Chuang FM, West BL, Baxter JD & Schaufele F 1997 Activities in Pit-1 determine whether receptor interacting protein 140 activates or inhibits Funding Pit-1/nuclear receptor transcriptional synergy. Molecular Endocrinology 11 1332–1341. Coulthard VH, Matsuda S & Heery DM 2003 An extended LXXLL motif This work was supported, in part, by a Health Sciences Research Grand to H sequence determines the nuclear receptor binding specificity of TRAP220. N, and a Grant-in Aid for Scientific Research to S S and H N from the Journal of Biological Chemistry 278 10942–10951. Ministry of Education, Culture, Sports, Science, and Technology in Japan. Crawford SE, Qi C, Misra P,Stellmach V,Rao MS, Engel JD, Zhu Y & Reddy JK 2002 Defects of the heart, eye, and in peroxisome Acknowledgements proliferator activator receptor-binding protein (PBP) null embryos implicate GATA family of transcription factors. Journal of Biological We are grateful to the following researchers for providing the plasmids; Drs Chemistry 277 3585–3592. Kazuhiko Umesono (Kyoto University, Japan), Ronald M Evans (The Salk Dasen JS, O’Connell SM, Flynn SE, Treier M, Gleiberman AS, Szeto DP, Institute, La Jolla, USA), Akihiro Sakurai (Shinshu University, Matsumoto, Hooshmand F, Aggarwal AK & Rosenfeld MG 1999 Reciprocal Japan), Masayuki Yamamoto (Tsukuba University, Tsukuba, Japan), Akira interactions of Pit1 and GATA2 mediate signaling gradient-induced Kakizuka (Kyoto University, Kyoto, Japan), Takashi Nagaya (Nagoya determination of pituitary cell types. Cell 97 587–598. University, Nagoya, Japan), Keita Tatsumi (Osaka University, Osaka, Japan), Ferreira R, Ohneda K, Yamamoto M & Philipsen S 2005 GATA1 function, a Mesut Muyan (University of Rochester Medical Center, Rochester, USA), paradigm for transcription factors in hematopoiesis. Molecular and Cellular and Leonard P Freedman (Memorial Sloan-Kettering Cancer Center, New Biology 25 1215–1227. York, USA). We also thank Dr P L Mellon (University of California, San Franklyn JA, Wood DF, Balfour NJ, Ramsden DB, Docherty K & Sheppard Diego, CA, USA) for providing TaT1 cells and Dr Keiichi Itoi (Tohoku MC 1987 Modulation by oestrogen of thyroid hormone effects on University, Japan) for the sequence information of the RT-PCR primers. thyrotrophin gene expression. Journal of Endocrinology 115 53–59. Furlanetto TW, Nguyen LQ & Jameson JL 1999 Estradiol increases proliferation and down-regulates the sodium/iodide symporter gene in FRTL-5 cells. Endocrinology 140 5705–5711. References Gittoes NJ, McCabe CJ, Verhaeg J, Sheppard MC & Franklyn JA 1997 Thyroid hormone and estrogen receptor expression in normal pituitary and Acevedo ML & Kraus WL 2003 Mediator and p300/CBP-steroid receptor nonfunctioning tumors of the anterior pituitary. Journal of Clinical coactivator complexes have distinct roles, but function synergistically, Endocrinology and Metabolism 82 1960–1967. during estrogen receptor a-dependent transcription with chromatin Gordon DF,Lewis SR, Haugen BR, James RA, McDermott MT, Wood WM templates. Molecular and Cellular Biology 23 335–348. & Ridgway EC 1997 Pit-1 and GATA-2 interact and functionally Alotaibi H, Yaman EC, Demirpence E & Tazebay UH 2006 Unliganded cooperate to activate the thyrotropin beta-subunit promoter. Journal of C K estrogen receptor-a activates transcription of the mammary gland Na /I Biological Chemistry 272 24339–24347. symporter gene. Biochemical and Biophysical Research Communications 345 Gordon DF, Woodmansee WW, Black JN, Dowding JM, Bendrick-Peart J, 1487–1496. Wood WM & Ridgway EC 2002 Domains of Pit-1 required for Arafah BM 2001 Increased need for thyroxine in women with hypothyroidism transcriptional synergy with GATA-2 on the TSH beta gene. Molecular and during estrogen therapy. New England Journal of Medicine 344 1743–1749. Cellular Endocrinology 196 53–66. www.endocrinology-journals.org Journal of Endocrinology (2008) 199, 113–125

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access 124 K NAGAYAMA and others . E2 inhibits TSHb gene expression

Gordon DF, Tucker EA, Tundwal K, Hall H, Wood WM & Ridgway EC Lisboa PC, Curty FH, Moreira RM, Oliveira KJ & Pazos-Moura CC 2001 2006 MED220/thyroid receptor-associated protein 220 functions as a Sex modulate rat anterior pituitary and liver iodothyronine transcriptional coactivator with Pit-1 and GATA-2 on the thyrotropin-beta deiodinase activities. Hormone and Metabolic Research 33 532–535. promoter in thyrotropes. Molecular Endocrinology 20 1073–1089. Liu J & Cui S 2005 Ontogeny of estrogen receptor (ER) alpha and its Gothe S, Wang Z, Ng L, Kindblom JM, Barros AC, Ohlsson C, Vennstrom B co-localization with pituitary hormones in the pituitary gland of chick & Forrest D 1999 Mice devoid of all known thyroid hormone receptors are embryos. Cell and Tissue Research 320 235–242. viable but exhibit disorders of the pituitary–thyroid axis, growth, and bone Maia AL, Harney JW & Larsen PR 1996 Is there a negative TRE in the maturation. Genes and Development 13 1329–1341. luciferase reporter cDNA? Thyroid 6 325–328. Gross HA, Appleman MD Jr & Nicoloff JT 1971 Effect of biologically active Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schutz G, Umesono K, steroids on thyroid function in man. Journal of Clinical Endocrinology and Blumberg B, Kastner P, Mark M, Chambon P et al. 1995 The nuclear Metabolism 33 242–248. receptor superfamily: the second decade. Cell 83 835–839. Haugen BR, McDermott MT, Gordon DF, Rupp CL, Wood WM & Mariotti S, Barbesino G, Caturegli P,Bartalena L, Sansoni P,Fagnoni F,Monti Ridgway EC 1996 Determinants of thyrotrope-specific thyrotropin beta D, Fagiolo U, Franceschi C & Pinchera A 1993 Complex alteration of promoter activation. Cooperation of Pit-1 with another factor. Journal of thyroid function in healthy centenarians. Journal of Clinical Endocrinology and Biological Chemistry 271 385–389. Metabolism 77 1130–1134. Hayakawa F, Towatari M, Ozawa Y, Tomita A, Privalsky ML & Saito H 2004 Matsushita A, Sasaki S, Kashiwabara Y, Nagayama K, Ohba K, Iwaki H, Functional regulation of GATA-2 by acetylation. Journal of Leukocyte Biology Misawa H, Ishizuka K & Nakamura H 2007 Essential role of GATA2 in the 75 529–540. negative regulation of thyrotropin beta gene by thyroid hormone and its Heck S, Kullmann M, Gast A, Ponta H, Rahmsdorf HJ, Herrlich P & receptors. Molecular Endocrinology 21 865–884. Cato AC 1994 A distinct modulating domain in glucocorticoid receptor McInerney EM, Weis KE, Sun J, Mosselman S & Katzenellenbogen BS 1998 monomers in the repression of activity of the transcription factor AP-1. Transcription activation by the human estrogen receptor subtype beta (ER EMBO Journal 13 4087–4095. beta) studied with ER beta and ER alpha receptor chimeras. Endocrinology Herrlich P 2001 Cross-talk between glucocorticoid receptor and AP-1. 139 4513–4522. 20 2465–2475. McKay LI & Cidlowski JA 1999 Molecular control of immune/inflammatory Hollowell JG, Staehling NW, Flanders WD, Hannon WH, Gunter EW, responses: interactions between nuclear factor-k B and steroid receptor- Spencer CA & Braverman LE 2002 Serum TSH, T(4), and thyroid signaling pathways. Endocrine Reviews 20 435–459. antibodies in the United States population (1988 to 1994): National Health Metivier R, Petit FG, Valotaire Y & Pakdel F 2000 Function of N-terminal and Nutrition Examination Survey (NHANES III). Journal of Clinical of the estrogen receptor requires a potential alpha- helical structure and is negatively regulated by the A domain. Molecular Endocrinology and Metabolism 87 489–499. Endocrinology 14 1849–1871. Ito M & Roeder RG 2001 The TRAP/SMCC/mediator complex and Metivier R, Penot G, Flouriot G & Pakdel F 2001 Synergism between function. Trends in Endocrinology and Metabolism ERalpha transactivation function 1 (AF-1) and AF-2 mediated by steroid 12 127–134. receptor coactivator protein-1: requirement for the AF-1 alpha-helical core Ito M, Yuan CX, Okano HJ, Darnell RB & Roeder RG 2000 Involvement and for a direct interaction between the N- and C-terminal domains. of the TRAP220 component of the TRAP/SMCC coactivator complex Molecular Endocrinology 15 1953–1970. in and thyroid hormone action. Molecular Cell 5 Nakano K, Matsushita A, Sasaki S, Misawa H, Nishiyama K, Kashiwabara Y & 683–693. Nakamura H 2004 Thyroid-hormone-dependent negative regulation of Kalaitzidis D & Gilmore TD 2005 Transcription factor cross-talk: the estrogen thyrotropin beta gene by thyroid hormone receptors: study with a new receptor and NF-kB. Trends in Endocrinology and Metabolism 16 46–52. experimental system using CV1 cells. Biochemical Journal 378 549–557. Kawai K, Sasaki S, Morita H, Ito T, Suzuki S, Misawa H & Nakamura H 2004 Nissen RM & Yamamoto KR 2000 The glucocorticoid receptor inhibits Unliganded thyroid hormone receptor-beta1 represses NFkB by interfering with serine-2 phosphorylation of the RNA alpha/oxysterol-dependent transactivation. Endocrinology 145 5515–5524. polymerase II carboxy-terminal domain. Genes and Development 14 Klinge CM, Bodenner DL, Desai D, Niles RM & Traish AM 1997 Binding of 2314–2329. type II nuclear receptors and estrogen receptor to full and half-site estrogen Norris JD, Fan D, Sherk A & McDonnell DP 2002 A negative coregulator for response elements in vitro. Nucleic Acids Research 25 1903–1912. the human ER. Molecular Endocrinology 16 459–468. Kobayashi Y, Kitamoto T, Masuhiro Y, Watanabe M, Kase T, Metzger D, Nowakowski BE & Maurer RA 1994 Multiple Pit-1-binding sites facilitate Yanagisawa J & Kato S 2000 p300 mediates functional synergism between estrogen responsiveness of the prolactin gene. Molecular Endocrinology 8 AF-1 and AF-2 of and beta by interacting directly 1742–1749. with the N-terminal A/B domains. Journal of Biological Chemistry 275 Ozawa Y, Towatari M, Tsuzuki S, Hayakawa F, Maeda T, Miyata Y, Tanimoto 15645–15651. M & Saito H 2001 3 associates with and represses the Kraus WL, McInerney EM & Katzenellenbogen BS 1995 Ligand- transcription factor GATA-2. Blood 98 2116–2123. dependent, transcriptionally productive association of the amino- and Pelletier G, Labrie C & Labrie F 2000 Localization of oestrogen receptor carboxyl-terminal regions of a steroid hormone nuclear receptor. PNAS alpha, oestrogen receptor beta and androgen receptors in the rat 92 12314–12318. reproductive organs. Journal of Endocrinology 165 359–370. Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S & Gustafsson JA 1996 Perissi V & Rosenfeld MG 2005 Controlling nuclear receptors: the circular Cloning of a novel receptor expressed in rat prostate and ovary. PNAS 93 logic of cofactor cycles. Nature Reviews. Molecular Cell Biology 6 542–554. 5925–5930. Pernasetti F, Caccavelli L, Van de Weerdt C, Martial JA & Muller M 1997 Lavinsky RM, Jepsen K, Heinzel T, Torchia J, Mullen TM, Schiff R, Del-Rio Thyroid hormone inhibits the human prolactin gene promoter by AL, Ricote M, Ngo S, Gemsch J et al. 1998 Diverse signaling pathways interfering with activating protein-1 and estrogen stimulations. Molecular modulate nuclear receptor recruitment of N-CoR and SMRT complexes. Endocrinology 11 986–996. PNAS 95 2920–2925. Sasaki S, Nakamura H, Tagami T, Miyoshi Y & Nakao K 1995 Functional Leong H, Sloan JR, Nash PD & Greene GL 2005 Recruitment of histone properties of a mutant T3 receptor beta (R338W) identified in a subject deacetylase 4 to the N-terminal region of estrogen receptor alpha. Molecular with pituitary resistance to thyroid hormone. Molecular and Cellular Endocrinology 19 2930–2942. Endocrinology 113 109–117. Lima LP, Barros IA, Lisboa PC, Araujo RL, Silva AC, Rosenthal D, Ferreira Sasaki S, Lesoon-Wood LA, Dey A, Kuwata T, Weintraub BD, Humphrey G, AC & Carvalho DP 2006 Estrogen effects on thyroid iodide uptake and Yang WM, Seto E, Yen PM, Howard BH et al. 1999 Ligand-induced thyroperoxidase activity in normal and ovariectomized rats. Steroids 71 recruitment of a histone deacetylase in the negative-feedback regulation of 653–659. the thyrotropin beta gene. EMBO Journal 18 5389–5398.

Journal of Endocrinology (2008) 199, 113–125 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access E2 inhibits TSHb gene expression . K NAGAYAMA and others 125

Schaufele F 1999 Regulation of estrogen receptor activation of the prolactin Vasudevan N, Ogawa S & Pfaff D 2002 Estrogen and thyroid hormone /promoter by antagonistic activation function-2-interacting receptor interactions: physiological flexibility by molecular specificity. proteins. Molecular Endocrinology 13 935–945. Physiological Reviews 82 923–944. Scott RE, Wu-Peng XS, YenPM, Chin WW & Pfaff DW 1997 Interactions of Wasco EC, Martinez E, Grant KS, St Germain EA, St Germain DL & estrogen- and thyroid hormone receptors on a Galton VA 2003 Determinants of iodothyronine deiodinase activities in estrogen response element (ERE) sequence: a comparison with the rodent uterus. Endocrinology 144 4253–4261. vitellogenin A2 consensus ERE. Molecular Endocrinology 11 1581–1592. Webb P, Nguyen P, Shinsako J, Anderson C, Feng W, Nguyen MP, Chen D, Scully KM, Gleiberman AS, Lindzey J, Lubahn DB, Korach KS & Rosenfeld Huang SM, Subramanian S, McKinerney E et al. 1998 Estrogen receptor MG 1997 Role of estrogen receptor-alpha in the anterior pituitary gland. activation function 1 works by binding p160 coactivator proteins. Molecular Molecular Endocrinology 11 674–681. Endocrinology 12 1605–1618. Sekulic M, Lovren M & Milosevic V 1998 Immunoreactive TSH cells in the Wondisford FE 1996 Factors that control thyroid function: thyrotropin. In pituitary of female middle-aged rats after treatment with estradiol or Werner and Ingbar’s The Thyroid Seventh Edition, pp 190–207. calcium. Acta Histochemica 100 185–191. Eds LE Braverman & RD Utiger. Philadelphia, USA: Lippincott-Raven Sjoberg M & Vennstrom B 1995 Ligand-dependent and -independent Publishers. transactivation by thyroid hormone receptor beta 2 is determined by the Wondisford FE, Farr EA, Radovick S, Steinfelder HJ, Moates JM, structure of the hormone response element. Molecular and Cellular Biology 15 McClaskey JH & Weintraub BD 1989 Thyroid hormone inhibition of 4718–4726. human thyrotropin beta-subunit gene expression is mediated by a Sosic-Jurjevic B, Filipovic B, Milosevic V, Nestorovic N, Manojlovic- cis-acting element located in the first . Journal of Biological Chemistry Stojanoski M, Brkic B & Sekulic M 2005 Chronic estradiol exposure 264 14601–14604. Wood WM, Ocran KW, Kao MY, Gordon DF, Alexander LM, Gutierrez- modulates thyroid structure and decreases T4 and T3 serum levels in middle-aged female rats. Hormone Research 63 48–54. Hartmann A & Ridgway EC 1990 Protein factors in thyrotropic tumor nuclear Stefaneanu L, Kovacs K, Horvath E, Lloyd RV,Buchfelder M, Fahlbusch R & extracts bind to a region of the mouse thyrotropin beta-subunit promoter Smyth H 1994 In situ hybridization study of estrogen receptor messenger essential for expression in thyrotropes. Molecular Endocrinology 4 1897–1904. ribonucleic acid in human adenohypophysial cells and pituitary adenomas. Yang Z & Privalsky ML 2001 Isoform-specific transcriptional regulation by Journal of Clinical Endocrinology and Metabolism 78 83–88. thyroid hormone receptors: hormone-independent activation operates Steger DJ, Hecht JH & Mellon PL 1994 GATA-binding proteins regulate the through a steroid receptor mode of co-activator interaction. Molecular human gonadotropin alpha-subunit gene in the placenta and pituitary Endocrinology 15 1170–1185. gland. Molecular and Cellular Biology 14 5592–5602. Yi P,Bhagat S, Hilf R, Bambara RA & Muyan M 2002 Differences in the abilities Stumpf M, Waskow C, Krotschel M, van Essen D, Rodriguez P, Zhang X, of estrogen receptors to integrate activation functions are critical for subtype- Guyot B, Roeder RG & Borggrefe T 2006 The mediator complex specific transcriptional responses. Molecular Endocrinology 16 1810–1827. functions as a coactivator for GATA-1 in via subunit Ying C, Lin DH, Sarkar DK & Chen TT 1999 Interaction between estrogen Med1/TRAP220. PNAS 103 18504–18509. receptor and Pit-1 protein is influenced by estrogen in pituitary cells. Surks MI & Hollowell JG 2007 Age-specific distribution of serum thyrotropin Journal of Steroid Biochemistry and Molecular Biology 68 145–152. and antithyroid antibodies in the US population: implications for the Yuan CX, Ito M, Fondell JD, Fu ZY & Roeder RG 1998 The TRAP220 prevalence of subclinical hypothyroidism. Journal of Clinical Endocrinology component of a thyroid hormone receptor- associated protein (TRAP) and Metabolism 92 4575–4582. coactivator complex interacts directly with nuclear receptors in a ligand- Surks MI & Sievert R 1995 Drugs and thyroid function. New England dependent fashion. PNAS 95 7939–7944. Journal of Medicine 333 1688–1694. Yusta B, Alarid ET, Gordon DF, Ridgway EC & Mellon PL 1998 The Tao Y, Williams-Skipp C & Scheinman RI 2001 Mapping of glucocorticoid thyrotropin beta-subunit gene is repressed by thyroid hormone in a novel thyrotrope cell line, mouse T alphaT1 cells. Endocrinology 139 4476–4482. receptor DNA binding domain surfaces contributing to of Zhang X, Jeyakumar M & Bagchi MK 1996 Ligand-dependent cross-talk NF-kappa B and induction of apoptosis. Journal of Biological Chemistry 276 between steroid and thyroid hormone receptors. Evidence for common 2329–2332. transcriptional coactivator(s). Journal of Biological Chemistry 271 14825–14833. Tillman JB, Crone DE, Kim HS, Sprung CN & Spindler SR 1993 Zhang X, Krutchinsky A, Fukuda A, Chen W, Yamamura S, Chait BT & Promoter independent down-regulation of the firefly luciferase gene by Roeder RG 2005 MED1/TRAP220 exists predominantly in a TRAP/ T and T receptor in CV1 cells. Molecular and Cellular Endocrinology 95 3 3 mediator subpopulation enriched in RNA polymerase II and is required for 101–109. ER-mediated transcription. Molecular Cell 19 89–100. Umesono K & Evans RM 1989 Determinants of target gene specificity for Zhu YS, Yen PM, Chin WW & Pfaff DW 1996 Estrogen and thyroid hormone steroid/thyroid hormone receptors. Cell 57 1139–1146. interaction on regulation of gene expression. PNAS 93 12587–12592. Vasudevan N, Koibuchi N, Chin WW & Pfaff DW 2001a Differential crosstalk between estrogen receptor (ER)alpha and ERbeta and the thyroid hormone receptor isoforms results in flexible regulation of the consensus ERE. Brain Research. Molecular Brain Research 95 9–17. Received in final form 11 July 2008 Vasudevan N, Zhu YS, Daniel S, Koibuchi N, Chin WW & Pfaff D 2001b Crosstalk between oestrogen receptors and thyroid hormone receptor Accepted 22 July 2008 isoforms results in differential regulation of the preproenkephalin gene. Made available online as an Accepted Preprint Journal of Neuroendocrinology 13 779–790. 23 July 2008

www.endocrinology-journals.org Journal of Endocrinology (2008) 199, 113–125

Downloaded from Bioscientifica.com at 09/23/2021 04:48:26PM via free access